Solved: Meta Ads CPA Increase Since March - Any Recommendations?

Hi there,

Thanks for reaching out! I had a look at the situation you described with your cosmetics brand's Meta ads. It’s a really common problem, so don't feel like you're alone in this. An increase in CPA, especially on a platform like Meta, is something we see quite a bit. The platform is always in flux, competition waxes and wanes, and what was working six months ago can suddenly just stop.

You mentioned you've been testing a lot of things, which is great, but it sounds like you're hitting a wall. I'm happy to give you some initial thoughts and guidance based on what I've seen work for other eCommerce brands. This won't be a magic bullet, but it should give you a new way to look at the problem and some concrete things to try.

We'll need to look at your audience targeting...

This is almost always the first place I start. You can have the best creatives in the world, but if you're showing them to the wrong people, or to the right people too many times, your costs will skyrocket. You mentioned changing creatives a lot, but what about the audiences? If you're hammering the same audience pools, even with fresh ads, you'll eventually hit a point of saturation and fatigue. It's like telling the same group of friends the same story but with slightly different words each time; eventually, they just tune out.

The core issue I often see is a flat approach to targetting. A lot of brands just lump everyone into one big bucket. A much better approach is to structure your campaigns based on the customer's journey and their relationship with your brand. Think of it as a funnel: Top (ToFu), Middle (MoFu), and Bottom (BoFu).

• -> BoFu (Bottom of Funnel): These are your warmest audiences. They know you, they've been on your site, maybe they've even added a product to their cart. These people should be your absolute top priority. They are the closest to converting and should, in theory, give you the lowest CPA. This includes people who have visited your checkout page, added to cart, or even previous customers. I remember one campaign we ran for a women's apparel brand where we saw a 691% return by really focusing on retargeting these warm audiences with specific offers.
• -> MoFu (Middle of Funnel): These folks are aware of you but aren't quite ready to buy. They might have watched one of your videos, engaged with an Instagram post, or visited a blog post. The goal here is to nurture them and move them down to the BoFu stage.
• -> ToFu (Top of Funnel): This is your cold traffic. People who have likely never heard of you. This is where you use interest-based targeting and Lookalike Audiences to find new potential customers. This is also where your CPA will naturally be the highest.

If your overall CPA is creeping up, it might be because you're spending too much at the top of the funnel, or your bottom-of-funnel audiences aren't being treated with the priority they deserve. You need to segment these audiences into seperate campaigns or ad sets so you can control the budget and see exactly where your money is perfoming best. It's a bit more work to set up, but it gives you so much more control.

Another thing to think about is the quality of your Lookalike Audiences. Are you creating them from a broad source like 'All Website Visitors'? Or are you getting more specific? A Lookalike built from your list of 'Highest Value Previous Customers' is going to be infinitly more powerful than one built from people who just glanced at your homepage. You need at least 100 people in a source audience, but honestly, you want a few thousand for it to be really effective. Prioritse creating Lookalikes from your best customers first.

Finally, let's get deeper than demographics. Who is your Ideal Customer Profile (ICP)? And I don't mean "women aged 25-45 in Europe". That's a demographic, not a profile. You need to understand their *nightmare*. What is the specific, urgent problem your cosmetic brand solves? Is it the frustration of dealing with adult acne? The insecurity of seeing the first signs of aging? The desire for a simple routine in a confusing market? When you define your customer by their *pain*, your ads stop being ads and start being solutions. Your ad copy becomes more potent, and your interest targeting becomes much sharper. Instead of targeting a broad interest like "Skincare", you can target followers of specific dermatologists, brands that solve similar problems, or magazines they actually read.

Here’s a rough idea of how you could structure and prioritise your audiences. The ones at the top should get the most attention and often, the best results.

I'd say your offer might be the problem...

This is the elephant in the room for so many businesses. You mentioned your product and offer haven't changed since you started. In the fast-moving world of online retail, and especially cosmetics, what worked in 2022 might be feeling a bit stale in mid-2024. Your competitors are constantly evolving their offers, and customer expectations shift.

The single biggest reason I see campaigns fail, even with brilliant ads and perfect targeting, is a weak offer. An offer that lacks a clear sense of urgency or high perceived value will always struggle. You are asking someone to stop scrolling through pictures of their friends and family, click a link, pull out their credit card, and trust a brand they might not know well. The offer has to be compelling enough to overcome all that friction.

It's not just about the product itself, but how you frame it. Are you just selling a single bottle of serum? Or are you selling a "Radiant Skin Starter Kit" that bundles your top 3 products for a 20% saving? Are you selling a one-off purchase, or a subscription that ensures they never run out and get a small discount for their loyalty? A simple shift in the offer can completely change your unit economics and advertising viability.

This brings me to a really important concept that many brands overlook: Customer Lifetime Value (LTV). Most businesses are fixated on the immediate Cost Per Acquisition (CPA). They see a €30 CPA and panic. But what if that €30 customer goes on to spend €200 with you over the next year? Suddenly, that €30 looks like a brilliant investment. If you don't know your LTV, you're flying blind. You can't possibly know what a "good" CPA is.

Let's run a quick, hypothetical caluclation. This is a simplified model, but it demonstrates the point:

• -> Average Revenue Per Account (ARPA): Let's say your average order value (AOV) is €60. For a subscription, this would be the monthly fee.
• -> Gross Margin %: After product costs, shipping, etc., what's your profit? Let's say it's 65%.
• -> Monthly Churn Rate: This is the tricky one for non-subscription businesses. Let's simplify and estimate the percentage of customers who *don't* make a repeat purchase within, say, 3 months. Let's say churn is high, around 20% per month on average for your customer base.

The LTV formula is: (ARPA * Gross Margin %) / Monthly Churn Rate

LTV = (€60 * 0.65) / 0.20
LTV = €39 / 0.20 = €195

In this completely hypothetical scenario, each customer you acquire is worth €195 in gross margin to your business over their lifetime. A healthy LTV to CAC (Customer Acquisition Cost) ratio is generally considered to be 3:1. This means you could afford to spend up to €65 (€195 / 3) to acquire a customer and still have a very healthy, scalable business. Suddenly your €25-€30 CPA doesn't look so scary, does it? It looks like an opportunity.

The whole point of this is to shift your focus from "How can I lower my CPA?" to "How can I increase my LTV so I can afford to pay more for a customer than my competition?". This is how you win. You can increase LTV by improving customer retention, launching new products for them to buy, or by creating a subscription/membership offer. A stronger offer not only converts better (lowering CPA) but also increases LTV, giving you a powerful double-whammy.

You'll need a solid testing framework...

You said you "test so much different things," which I hear a lot. But often this testing is unstructured. It's like throwing spaghetti at the wall to see what sticks. It's chaotic, it's hard to learn from, and it burns through your budget. You need a systematic proccess.

First, your campaign structure should reflect your funnel. Don't jumble everything together. Have seperate, long-term campaigns for each stage:

1. A Prospecting Campaign (ToFu): This is where you test your cold audiences – your interests and Lookalikes. The goal here is to find new, scalable audiences that bring in customers at an acceptable CPA.
2. A Retargeting Campaign (MoFu/BoFu): This campaign is purely for your warm audiences – website visitors, cart abandoners, video viewers, etc. The ad copy here should be different. It should address their potential objections, offer social proof (reviews!), or create urgency ("Your cart expires soon!").
3. A Retention/Cross-sell Campaign (Existing Customers): This is for people who have already bought from you. Show them new products, offer them a special "loyal customer" discount. This is often the most profitable campaign you can run.

Within each campaign, you test audiences at the ad set level. And within each ad set, you test your creatives at the ad level. This seperation is critical. It lets you isolate variables. If an ad set is failing, you know it's the audience. If an ad is failing, you know it's the creative or copy. You need a strict rule for making decisions. For example, if an ad set has spent 2x your target CPA with no purchases, it gets turned off. No emotion, just data.

You also mentioned you are active in all countries. Are you running these in the same campaign? This could be a huge source of your problem. The CPA in Germany is going to be wildly different from the CPA in Portugal or Poland. Consumer behaviour, competition, and purchasing power vary massively. You should definitly be running seperate campaigns (or at least ad sets) for different countries or country groups (e.g., 'DACH region', 'Nordics', 'Southern Europe'). Lumping them all together means the algorithm will likely spend the budget in the cheapest-to-reach country, which might not be where your best customers are. This can really skew your overall CPA and hide strong performance in one country behind weak performance in another.

Finally, let's talk about creatives again. Instead of just "changing" them, think about what you're testing. Use the Before-After-Bridge framework for your copy. For a cosmetic brand, your creative tests should be relentless. You should be testing:

• -> Formats: Static Image vs. Carousel vs. Video vs. UGC (User-Generated Content). For cosmetics, UGC is gold. A real customer showing their results in a simple phone video is often far more powerful than a slick, professional studio shot. We have SaaS clients seeing amazing results with UGC, so for a visual product like yours, it's a must-test.
• -> Hooks: The first 3 seconds of your video or the headline of your ad. Test different pain points. "Tired of complicated skincare routines?" vs. "The secret to glowing skin in 30 days".
• -> Angles: Is your product about "clean ingredients"? "Scientifically proven results"? "Affordable luxury"? Each of these is a different angle that will resonate with a different sub-segment of your audience. Test them as distinct concepts.

This structured approach transforms testing from a gamble into a process of discovery. You're no longer just hoping to get lucky; you're systematically finding what works and cutting what doesn't.

I know this is a lot to take in, but a rising CPA is usually a symptom of a deeper strategic issue rather than a simple platform glitch. It’s often a combination of audience fatigue, a stale offer, and an unstructured approach to testing.

I've detailed my main recommendations for you below:

Implementing a full strategic overhaul like this can be a daunting task, especially when you're also trying to run the rest of your business. It requires deep expertise, constant monitoring, and a disciplined approach. It's what we do all day, every day.